Why? If a certain race is shorter than others, and you have measured the Indivudual of that race standing before you to be 6′8″, are you supposed to downgrade your measurement?
I don’t understand this because you either made a typo or are leaving out information (that, for instance, in the hypothetical intelligence is negatively correlated with height).
Only if the correlation between race and bad things is eliminated once you take the other factor into account should you then stop using race
If the “bad things” and the “other factor” are identical, the correlation between race and bad things conditional on the other factor is zero, so you actually don’t have a reason to use race. The correlation between race and being shorter than 6′8“ is eliminated once you condition on being 6′8”.
But this doesn’t work for “bad things” such as “will X happen in the future”, since you can’t directly observe them until it’s too late to make any decision about them, racial or non-racial.
I was referring to standard racist ideas such as “people of that race are more likely to rob you”. If you can observe whether someone has already robbed you, race is irrelevant for determining if they have robbed you, but that only applies to the past. You can’t observe whether someone is going to rob you in the future the same way you can observe whether someone is 6′8″.
(Of course you can have other objections. My recent comments about not hiring people based on IQ apply here too:if you refuse to hire people of some race because they have a higher chance of robbing you, the same people will find themselves constantly not hired, and this is bad. I’m not arguing for racism; I’m just pointing out that this objection doesn’t hold up.)
Why? If a certain race is shorter than others, and you have measured the Indivudual of that race standing before you to be 6′8″, are you supposed to downgrade your measurement?
I don’t understand this because you either made a typo or are leaving out information (that, for instance, in the hypothetical intelligence is negatively correlated with height).
Corrected
Okay. I did say:
If the “bad things” and the “other factor” are identical, the correlation between race and bad things conditional on the other factor is zero, so you actually don’t have a reason to use race. The correlation between race and being shorter than 6′8“ is eliminated once you condition on being 6′8”.
But this doesn’t work for “bad things” such as “will X happen in the future”, since you can’t directly observe them until it’s too late to make any decision about them, racial or non-racial.
Do you have concrete examples? AFAICS we don’t have good ways of predicting future behaviour in general.
I was referring to standard racist ideas such as “people of that race are more likely to rob you”. If you can observe whether someone has already robbed you, race is irrelevant for determining if they have robbed you, but that only applies to the past. You can’t observe whether someone is going to rob you in the future the same way you can observe whether someone is 6′8″.
(Of course you can have other objections. My recent comments about not hiring people based on IQ apply here too:if you refuse to hire people of some race because they have a higher chance of robbing you, the same people will find themselves constantly not hired, and this is bad. I’m not arguing for racism; I’m just pointing out that this objection doesn’t hold up.)
An problem with racism is still confusion of correlation and causation....